Datasets:

sltttt
/

Text2Loc

This repository provides the dataset and auxiliary resources for Text2Loc_v2, associated with the paper Text2Loc++: Generalizing 3D Point Cloud Localization from Natural Language.

The released files are intended to support research on text-based 3D point cloud localization, including data loading, model evaluation, and reproduction of the main experimental setup.

Repository Structure

.
├── README.md
├── .gitattributes
├── data.zip
└── model_checkpoints.zip

File Description

`data.zip`

This archive contains the main dataset used for Text2Loc_v2.

It includes the processed data required for training and evaluation. Users who would like to reproduce the main results or train models on the full dataset should use this file.

Please unzip this archive before running the training or evaluation scripts.

`model_checkpoints.zip`

This archive contains the released model checkpoints.

These checkpoints can be used for evaluation, inference, or as initialization for further research. Please refer to the main GitHub repository for the expected checkpoint paths and loading instructions.

Recommended Usage

For regular use, please download and extract:

data.zip
model_checkpoints.zip

Please make sure the extracted file paths match the configuration used in the GitHub repository.

Notes

Additional lightweight files may be present in this repository for internal review or demonstration purposes. They are not required for normal usage, formal training, evaluation, or result reproduction.

License

This dataset and the associated resources are released under the Apache 2.0 License.

Citation

If you find this work useful in your research, please consider citing:

@inproceedings{xia2024text2loc,
  title     = {Text2Loc: 3D Point Cloud Localization from Natural Language},
  author    = {Xia, Yan and Shi, Letian and Ding, Zifeng and Henriques, Jo\~{a}o F. and Cremers, Daniel},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  pages     = {14958--14967},
  year      = {2024}
}

@article{xia2025text2locpp,
  title   = {Text2Loc++: Generalizing 3D Point Cloud Localization from Natural Language},
  author  = {Xia, Yan and Shi, Letian and Di, Yilin and Henriques, Jo\~{a}o F. and Cremers, Daniel},
  journal = {arXiv preprint arXiv:2511.15308},
  year    = {2025}
}

license: apache-2.0

Text2Loc_v2 Dataset and Resources

Links